Calorimeter Cable Assignments for
Front End Electronics and Trigger Systems
ZEUS NOTE NUMBER 90-058
C. Foudas, C. Fordham, J. Lackey, P. Robl
S. Silverstein, W. H. Smith
University of Wisconsin, Madison, WI 53706.

Abstract

In this Note we present the assignments of the ZEUS calorimeter Photomultiplier (PMT) signals to the inputs of the analog Front End Cards (FEC). Since a large portion of the analog sums required for the Calorimeter First Level Trigger (CFLT)[1] is formed on the FECs the assignment of FEC to PMTs is directly related to the CFLT supertower configuration. The analog signals from the Trigger outputs of the FECs are directed to the Trigger Sum Cards (TSC) where signals from left and right PMTs are summed. The last step in this summation process occurs at the front end of the Trigger Encoder Card (TEC) where all the signals that form a supertower are summed. The design of the TEC allows only two signals to be summed at its front end. Therefore, all other analog sums need to be formed on the ZEUS detector calorimeter by appropriately combining the PMTs on FECs and TSCs. We have used the ZEUS CFLT Monte Carlo Simulation[2] to form the CFLT supertower combinations. The PMT, FEC, TSC and TEC tables along with their relationships have been entered in ADAMO[3].

Introduction

Due to the non-projective geometry of the ZEUS calorimeter[4], the path taken by a particle from the interaction point often travels through the EMC of one tower and the HAC of an adjacent tower. While ``projective'' combinations of the EMC and HAC sections of different non-projective towers can be made with software in the SLT, TLT,[5] and offline, they must be made directly in the hardware of the FLT.

The creation of a projective geometry from the non-projective mechanical calorimeter geometry is complex. The tower geometry is set by locations of the EMC towers (groups of either four 5 x 20 cm or two 10 x 20 cm sections). The particles that traverse these then ``shadow'' various combinations of HAC towers. The first and 14th BCAL EMC towers shadow the outer ring of the FCAL and RCAL HAC towers, respectively, and do not shadow HAC towers in the BCAL. The top and bottom of the central FCAL and RCAL modules shadow HAC towers above and below them, respectively. The outer 6 or so FCAL and RCAL modules have EMC towers that shadow HAC towers in the adjacent modules. In many of these cases, an individual 20 x 20 cm EMC tower shadows more than 1 (in some cases, as many as 6) HAC towers.

The importance of projectivity is shown by the isolated electron trigger. In order to distinguish an electron from a hadron, the energy in the HAC that would be traversed by a particle that penetrates a certain EMC section must be much smaller than the energy in the EMC section. For many sections of the calorimeter (especially the FCAL and RCAL for theta greater than 20 degrees) a particle that traverses an EMC section of a physical tower does not penetrate the HAC section, but that of an adjacent tower. Therefore, it is inappropriate to use the energy in the HAC section to identify electrons that deposited energy in the EMC section of the same mechanical tower. This check of HAC/EMC energy is done every 96 nsec for each tower in the calorimeter. The two energies to be compared, HAC and EMC, must be fed to the same memory lookup table in order to perform this test. In order to achieve this, the hardware requires that the EMC and HAC pulseheights to be compared be digitized on adjacent electronics channels.

The CAL FLT electronics is constrained to have ``projective'' combinations of EMC and HAC towers cabled to the TEC front end. The pulseheight digitization is done in the CAL FLT Trigger Encoder Cards in the Rucksack. The PMT signals are first brought to the Frond End Cards. In the FCAL four trigger sums are formed from the 12 left or 12 right PMT signals on each FEC. The 4 trigger sums from one right PMT FEC and its corresponding left PMT FEC are combined to form 4 (or less) total trigger sums on each Trigger Sum Card (TSC). On the TSC the analog signal as are further combined to form the CFLT super towers. The total trigger sums from the Trigger Sum Cards are transmitted to the Trigger Encoder Cards in the Rucksack. The Trigger Encoder Cards are able to sum two separate inputs for each digitization channel.

PMT to FEC assignments for the FCAL and BCAL

The Front End Cards designed for the ZEUS FCAL and BCAL have 12 PMT input connectors and 4 CFLT Sum Output connectors. This type of FEC is defined as a type ``A'' FCAL/BCAL FEC. A description of the 16 FEC connectors is given in Table 1.

BCAL/FCAL Front End Card Connectors
Top View From Right To Left (Figure 1)
CONNECTOR NUMBER	DESCRIPTION
1	Trigger Sum Output #1
2-5	PMT Inputs #1-#4
6	Trigger Sum Output #2
7-10	PMT Inputs #5-#8
11	Trigger Sum Output #3
12-15	PMT Inputs #9-#12
16	Trigger Sum Output #4

Table 1. : The FCAL abd BCAL FEC Connectors.

The four Trigger Sum Outputs (TSO) are defined as the analog sums of PMT input signals PMT(1-12):

$$\begin{array}{l} TSO \char93 1 = PMT(1) + PMT(2) + PMT(3) + P... ...\char93 4 = PMT(9) + PMT(10) + PMT(11) + PMT(12)\\ \end{array}$$ (1)

In the case that this summation configuration is not adequate the FEC has to be modified according to the specific needs. Modifying an FCAL/BCAL FEC requires disassembling it and rerouting the analog sums to the various TSOs. Therefore one important constraint in assigning PMT signals to FECs was to avoid having many different types of FCAL/BCAL FECs. As we shall see later only a small number of one extra type of FEC had to be introduced for the FCAL.

The PMT to FEC assignments were made with the following rules:

• Minimize the number of required FECs.
• Minimize the number of required TSCs.
• Minimize the number of modified FECs.
• Since FECs are positioned in the dataways on each module, input signals of each FEC must come from the module on which the given FEC resides. Furthermore, since the majority of the FCAL modules have top and bottom dataways, signals coming from PMTs at the bottom of the module in general cannot be assigned to FEC located at the top of the module and vice versa.
• Left and right PMT signals should never mix on the same FEC.
• An FEC connected to the PMT signals from the left side of a tower or towers should be configured identically with its counterpart which is connected to the PMT signals from the right side of that tower.
• Assign sums according to the CFLT configuration[1].
• Region boundaries of the Second level Trigger should not be crossed by half FECs[5].
• Assignments of PMTs to FEC should take in to account the length of the PMT cables[7].
• Leave several free half FECs and free dataway slots for the Laser system.

In order to satisfy all of these constraints in the FCAL we had to introduce one more type of FEC, called type ``B''. The configuration of the type ``B'' FEC is shown in Table 1. The four Trigger Sum Outputs (TSO) for the type ``B'' are defined as the analog sums of PMT input signals PMT(1-12) :

$$\begin{array}{l} TSO \char93 1 = PMT(1) + PMT(2) + PMT(3) + P... ...\char93 4 = PMT(9) + PMT(10) + PMT(11) + PMT(12)\\ \end{array}$$ (2)

A picture of a modified type ``B'' Front End Card printed circuit board is shown in Fig. 1.

We have allocated:

• 396 FECs for the FCAL PMTs out of which 388 are of type ``A'' and 8 of type ``B''.
• 448 FECs for the BCAL PMTs all of which are of type ``A''.

Individual pictures that illustrate the connections of FEC to PMT are shown in Figures 2-24 for the FCAL and Figure 26 for the BCAL. In each figure we show a one-sided picture of an FCAL module with solid dots indicating left or right PMTs. Horizontal lines show the FEC boundaries. Identical contiguous patterns fill towers that are summed in the CFLT trigger towers. Continuous lines between PMTs (solid dots) indicate sums formed on the FECs. In general, any PMTs belonging to the same pattern should be connected with a continuous line indicating that their signals are summed together on the FEC. Various limitations such as limited number of TSOs of the FEC, SLT region boundaries and others made it impossible to do that. Those cases have been studied individually and the summations will be done on the Trigger Sum Card (TSC). The TSC is designed to sum the left and right signals coming from the TSOs of the FECs. Via jumpers, it can also combine several signals coming from different towers. The physical locations of the FECs in the FCAL are shown in Fig 25 (a), 25(b). In Table 2 we show the locations of free half FECs and dataway slots for the FCAL. These FECs are located in the dataway slots furthest from the beam pipe.

FCAL FREE HALF FECs AND FREE DATAWAY SLOTS
MODULE	NUMBER OF FREE 1/2 FECS	NUMBER OF FREE FEC SLOTS
23	0	0 UP + 0 DOWN
22	0	2 UP + 2 DOWN
21	2 DOWN	0
20	0	0 UP + 0 DOWN
19	0	0 UP + 0 DOWN
18	0	0 UP + 0 DOWN
17	2 UP	0 UP + 1 DOWN
16	2 UP	2 UP + 0 DOWN
15	2 UP	2 UP + 0 DOWN
14	2 UP	2 UP + 0 DOWN
13	2 UP	2 UP + 0 DOWN
12	0	2 UP + 2 DOWN
11	2 DOWN	0 UP + 2 DOWN
10	2 DOWN	0 UP + 2 DOWN
09	2 DOWN	0 UP + 2 DOWN
08	2 DOWN	0 UP + 2 DOWN
07	2 DOWN	0 UP + 1 DOWN
06	0	0 UP + 0 DOWN
05	0	0 UP + 0 DOWN
04	1 UP	0 UP + 0 DOWN
03	2 UP	0
02	0	2 UP + 2 DOWN
01	0	0 UP + 0 DOWN

Table 2. : Free half FECs and empty data way slots in the FCAL.

PMT to FEC assignments for the RCAL

The Front End Cards designed for the ZEUS RCAL have 12 PMT input connectors and 8 Trigger Sum Output (TSO) connectors. The eight connectors closer to the edge of the card are the Trigger Sum Outputs. The next row of twelve connectors are the PMT input signals. The positions of PMT input number one and TSO number one are shown in Figure 27, which depicts the printed circuit board for the RCAL FECs. The default (no jumpers) type of RCAL FEC is defined as a type ``A'' RCAL FEC. The eight Trigger Sum Outputs (TSO) are defined as the analog sums of PMT input signals PMT(1-12):

$$\begin{array}{l} TSO \char93 1 = PMT(1) + PMT(2)\\ TSO \char... ...O \char93 7 = PMT(10)\\ TSO \char93 8 = PMT(12)\\ \end{array}$$ (3)

Note that inputs #5 and #11 are not connected to a trigger output in the default configuration of the RCAL FEC card. In addition, due to the design of the RCAL FEC printed circuit board, TSO #1 is always the sum of PMT inputs 1, 2 and TSO #5 is always the sum of PMT inputs 7, 8 (see Fig. 27).

This form of the RCAL FEC was not adequate for most of the RCAL, because most of the RCAL towers are made up of two EMC sections followed by one HAC section. Hence most of the FECs will be modified to reflect the need for equal numbers of sums of two inputs and unsummed inputs. These FECs are called RCAL type ``B''. For the RCAL type ``B'' FEC, the eight Trigger Sum Outputs (TSO) are defined as the analog sums of PMT input signals PMT(1-12) :

$$\begin{array}{l} TSO \char93 1 = PMT(1) + PMT(2)\\ TSO \char... ...7 = PMT(10) + PMT(11)\\ TSO \char93 8 = PMT(12)\\ \end{array}$$ (4)

This configuration can be produced by inserting two jumpers J1 and J2 as shown in Fig. 28. Note that this configuration, unlike the unjumpered configuration, allows the use of inputs 5 and 11 in the trigger.

A third type of RCAL FEC which makes larger sums is used in a few cases for convenience. This FEC, called RCAL type ``C'', is used only on the far left and far right modules of the RCAL, where many HAC sections have to be summed for assignment to the BCAL EMC trigger towers which shadow them. This type of FEC sums inputs in groups of four and two to give four outputs. The alterations are made as shown in Figure 29, by inserting jumpers J1, J2, J3, J4 between the trigger outputs on the cards. The assignments of the Trigger Sum Outputs (TSOs) are as follows:

$$\begin{array}{l} TSO \char93 1 = PMT(1) + PMT(2)\\ TSO \char... ...\char93 6 = PMT(9) + PMT(10) + PMT(11) + PMT(12)\\ \end{array}$$ (5)

Trigger Sum Outputs #3, #4, #7, and #8 are not used for the RCAL type ``C'' FEC.

We have allocated 222 FECs for the RCAL PMTs. These 222 FEC fall into three types as follows:

1.

36 are of RCAL type ``A''.

2.

178 are of RCAL type ``B''.

3.

8 are of RCAL type ``C''.

The assignments of FECs for each of the RCAL modules are shown in figures 30-52. In these figures, the HAC0/EMC sections are shown on the left. In addition to the information shown for the FCAL drawings, these drawings also show the Second Level Trigger (SLT) boundaries and the analog card types, with ``normal'' representing type A, ``(2-1)'' representing type B, and ``(4-1)'' representing type C. Some HAC0 cells are labelled with E or H; the E indicates that this cell is being treated as an electromagnetic cell for trigger purposes, while the H labels have been added to prevent confusion. The locations of the FECs in the dataways are shown in Fig 53 (a) and (b). In Table 3 we show the locations and the numbers of free half FECs and free dataway slots for the RCAL.

RCAL FREE HALF FECs AND FREE DATAWAY SLOTS
MODULE	NUMBER OF FREE 1/2 FECS	NUMBER OF FREE FEC SLOTS
23	0	2 UP + 0 DOWN
22	2 DOWN	0
21	2 DOWN	0
20	2 UP	2 UP + 0 DOWN
19	0	2 UP + 0 DOWN
18	0	2 UP + 0 DOWN
17	0	2 UP + 0 DOWN
16	2 UP	0
15	2 UP	0
14	2 UP	0
13	0	0
12	0	0
11	0	0
10	2 DOWN	0
09	2 DOWN	0
08	2 DOWN	0
07	0	0 UP + 2 DOWN
06	0	0 UP + 2 DOWN
05	0	0 UP + 2 DOWN
04	2 DOWN	0 UP + 2 DOWN
03	2 UP	0
02	2 UP	0
01	0	2 UP + 0 DOWN

Table 3. : Free half FECs and free data way slots in RCAL. All free half FECs, except in modules 3, 4, 20, and 21, are located in the slot furthest away from the beam pipe.

Using ADAMO to store the FEC Information

In order to make the PMT to FEC assignment information easily accessible to the detector assembly/installation and the CFLT/CSLT groups, for debugging their electronics, we have entered it into the ADAMO[3] data base. We have used ADAMO to store all the PMT, FEC, TSC, TEC assignments. Two of these tables, the PMT and the TEC table, are already implemented into the CFLT Monte Carlo simulation[1]. These tables were used as constraints (along with all the other constraints mentioned before) in order to fix the assignments of the FECs and TSCs.

Entering this information into ADAMO was a three step process. First, using an interactive Fortan program, all the information of the summation configurations shown in Figures 2-24, 26, and 30-52 was recorded into ASCII files. Second, the ASCII files from all the modules of the calorimeter where combined and the information was entered in to ADAMO. At the end, the contents of ADAMO were printed and compared by physicists with the original information in Fig 2-24, 26, and 30-52 in order to avoid any errors.

In Appendix A we give the ADAMO tables for the PMT to FEC assignments for the ZEUS calorimeter. The ``Feclist'' table contains all the information needed to connect the PMT signals to FECs. Each row in this table shows the configuration of a particular FEC somewhere in the ZEUS calorimeter and has 14 columns described below:

1.

Column 1 (CAL) gives the Calorimeter section information (FCAL, BCAL, RCAL)

2.

Column 2 (Number) gives the module number (1-23), followed by the FEC number within a module, and the FEC type (``A'', ``B'',``C''). For FCAL and BCAL we have counted the FECs from the bottom to the top of each module. For the BCAL modules FEC 1 and 2 are located in the pocket closest to the FCAL. In this numbering scheme[8], modules 1-FCAL and 1-RCAL are at negative x direction while 23-FCAL and 23-RCAL are at positive x direction. The positive y direction is from bottom to the top and the proton direction defines the positive z direction.

3.

Columns 3-14 (I(1-12)) represent the FEC inputs and their contents indicate the location of the PMT whose signal is connected to that input. The format of the contents of each column starts with the tower number (1-23) followed by the Cell name (E1-E4 or H0-2), and a character showing whether it is the right (R) or the left (L) PMT connected to the input. If the contents of I(1-12) are XXXXXXX then that input must be left disconnected. For FCAL/RCAL we have counted towers from the bottom to the top (positive y direction) and for BCAL from FCAL to RCAL.

FEC to TSC Assignments

The Trigger Sum Cards (TSCs) for the Calorimeter First Level Trigger have been described already[1]. The TSCs take as their inputs four pairs of left and right FEC outputs. The left and right inputs for each pair of FECs are summed together and sent to four outputs. If further sums are desired jumpers can be set on the cards so that various combinations of the four pairs of inputs can be summed. The summing configurations of the TSCs are the same for all regions of the calorimeter, although there are slight variations in the physical size of the cards between regions and also some adaption of the gain to suit the regions in which they will be used.

For the default TSC, type ``A'', the four Trigger Sum Card Outputs (TSCO) are defined as follows:

$$\begin{array}{l} TSCO \char93 1 = TSCI_{left}(1) + TSCI_{righ... ...SCO \char93 4 = TSCI_{left}(4) + TSCI_{right}(4)\\ \end{array}$$ (6)

where TSCI is the TSC input. The FEC outputs connected as a left and right pair to the TSC must correspond to the same cell or sum of cells; however, different pairs of inputs are not constrained to come from the same FECs.

Not all of the necessary sums of cells could be made on the FECs; therefore, the standard configuration of TSC had to be modified to allow for making sums of more than one pair of inputs from the FECs. Three such altered TSCs have been used: one which sums the third and fouth pairs of inputs to give a sum of two pairs of left and right FEC outputs (type ``B''), one which gives a sum of three pairs of FEC outputs (type ``C''), and one which gives two sums of two pairs of FEC outputs (type ``D''). These modified types are made by inserting input jumpers and removing output jumpers.

For TSC type ``B'', the four Trigger Sum Card Outputs (TSCO) are defined as follows:

$$\begin{array}{l} TSCO \char93 1 = TSCI_{left}(1) + TSCI_{righ... ...TSCI_{right}(4)\\ TSCO \char93 4 = disconnected\\ \end{array}$$ (7)

For TSC type ``C'', the four Trigger Sum Card Outputs (TSCO) are defined by:

$$\begin{array}{lll} TSCO \char93 1 & = & TSCI_{left}(1) + TSCI... ...isconnected\\ TSCO \char93 4 & = & disconnected\\ \end{array}$$ (8)

Lastly, TSC type ``D'' is jumpered to give the following outputs:

$$\begin{array}{l} TSCO \char93 1 = TSCI_{left}(1) + TSCI_{righ... ...CI_{right}(3) + TSCI_{left}(4) + TSCI_{right}(4)\\ \end{array}$$ (9)

The following considerations were used in assigning FEC outputs to TSCs:

• FECs can only be connected to TSCs on the same module.
• All sums of cells within the same module must be complete at the level of the TSCs; the TEC front end summation is only to be used for sums across module boundaries, since there is no other way to do the summation. Use of the TEC front end summation is not favored owing to the fact that the long cable run from the calorimeter module to the Rucksack is more likely to make the analog signals out of time with each other.
• The distance from TSC to FEC cannot be longer than the cable between them, which is 1.8 (1.9 for EMC) meters in the FCAL, 1.2 meters in the RCAL, and about 25 cm in the BCAL. This means that TSCs in the BCAL must be hooked to the FECs in the same pocket. Also, TSCs in the RCAL must be connected to FECs on the neighboring dataway, since the cables are not long enough to reach to the dataways on the other end of the module.
• The left and right sides of a cell (or sum of cells) should always be connected to the same pair of inputs on the TSC.
• Sums should be assigned according to the CFLT configuration[1].
• The configuration of trigger towers on the outputs of the TSCs must be such that splitting of the 8-pair cables from the TSCs to the TECs can be kept to a minimum.

The TSC assignments to the dataways are also shown in Figures 53 and 25; generally, they are assigned with TSC number increasing up the module. The numbers of TSCs of the different types are:

1.

Type ``A'': 160 in the BCAL, 90 in the RCAL, and 98 in the FCAL.

2.

Type ``B'': 64 in the BCAL, 98 in the RCAL, and 76 in the FCAL.

3.

Type ``C'': 13 in the RCAL, 2 in the FCAL.

4.

Type ``D'': 22 in the FCAL.

Using ADAMO to store the TSC Information

The FEC to TSC assignment information, like that for the PMT to FEC assignments, has been entered into the ADAMO data base. Again, this was done using a three step process, with an interactive Fortran program used to enter the information into an ASCII file for each module. Only the analog cards serving the left PMTs were entered into this ASCII file; the program then automatically made the same entry for the analog card serving the right PMT. Then these ASCII files for each module were combined and the information was read in by another fortran program which loaded it into ADAMO tables. The tables were printed and checked by physicists as well as being subjected to ADAMO checking. Appendix B gives these ADAMO tables for the FEC to TSC assignments. The ``TSClist'' table contains all the information needed to connect the FEC signals to the TSCs. Each line in the table corresponds to one TSC somewhere in the calorimeter. There are thirteen columns in the table; these are described below.

1.

Column 1 (CAL) gives the Calorimeter section information (FCAL, BCAL, RCAL).

2.

Column 2 (Modul) gives the module number. In the FCAL and RCAL, the modules are numbered 1 to 23 starting at negative x; in the BCAL, the modules are numbered 1 to 32 in increasing order in $\phi$ starting at $\phi=0$.

3.

Column 3 (TSCnu) gives the TSC number. These are numbered in increasing order starting with 1 in the bottom of the bottom dataway in the FCAL and RCAL, and starting in the pockets nearest the FCAL for the BCAL TSCs.

4.

Column 4 (Type) gives the TSC type; these are A, B, C, or D as described above.

5.

Column 5 (Dway) gives the dataway in which the TSC resides; these are either TOP or BOTT(bottom) for the RCAL and FCAL; the BCAL does not have dataways and the field is filled with ``XXXX''.

6.

Columns 6-13 (I(1-8)) represent the TSC inputs; consecutive pairs are for the left and right signals from one cell or group of cells. The content of these columns give the FEC output which is connected to that input. This information takes the form of a character string AA-B-C-D, where AA gives the FEC number in that module, B gives the output number on that FEC, C gives the FEC type, and D tells whether the signal is from the left or right side of a cell or group of cells. The left and right signals should come from FECs with consecutive numbers and from the same output on both FECs. If the contents of a column are XXXXXXXX then that input of the TSC is left disconnected.

TSC to Cable Connections

From the TSCs on the calorimeter modules eight-twisted-pair cables carry the analog signals to the TECs in the Rucksack. Because each TSC has four outputs, two TSCs can be connected to one cable. There is some constraint on the configuration of this connection; the cable harnesses require that any given TSC be attached to half of the cable and that the order of the channels on the cable strands be given by TSCO#4 to the first of the four cable strands, TSCO#3 to the second, TSCO#2 to the third, and TSCO#1 to the last of the set of four strands. In the FCAL and RCAL, the cable halves can be connected to any TSC on the same dataway as the fanout card; however, in the BCAL, the TSC to cable configuration is fixed completely by the cable harness.

Where freedom to determine the cable connections existed, the first priority was to minimize the amount of cable splitting necessary. This meant that signals going to the same place were kept on the same cable if at all possible. Splitting between TECs in one half of a crate was preferred over splitting over the whole crate; likewise, splitting between halves of a crate was preferred over splitting between regions. (One CFLT crate serves one CFLT region.)

Using ADAMO to store the Cable Assignments

The TSC to cable assignments have also been entered into the ADAMO database. This was done by writing an ASCII file with the pertinent information and then using a fortran program to read it in and enter the assignments into an ADAMO table. This table is shown in Appendix C. In the ``CableList'' table, each line represents a cable. There are twelve columns in the cable. The first three, CAL, Modul, and Cabl, form the label given on the cables: Cal-Mod-#. An explanation of the meaning of the columns is given below.

1.

Column 1 (CAL) gives the calorimeter section (FCAL, BCAL, or RCAL) from which the cable comes.

2.

Column 2 (Modul) gives the module from which the cable comes. The module numbering is the same as that used for the TSCs and FECs.

3.

Column 3 (Cabl) gives the cable number. Cables are numbered within a module from 1 to 7. Number 1 is always the control cable, number 2 is always a veto cable. In the FCAL and RCAL, if there is more than one veto cable coming from a module, it is given number three. The cables carrying the analog data signals from the TSCs to the TECs are numbered in increasing order within a dataway, starting with number 4, and also labeled by ``B'' for the bottom dataway or ``T'' for the top dataway. Thus, the data cables can go from 4T to 6T and 4B to 6B within one RCAL or FCAL module. If there is only one TSC dataway on a module, the data cables are numbered from 4B to 6B. In the BCAL, the cables are numbered 3-6 with 3 servicing the TSC nearest the FCAL and the consecutive numbers servicing consecutive pairs.

4.

Column 4 (Type) gives the type of the cable. The possible types are ``CONT'' for control, ``VETO'', and ``DATA''. These must correspond with the numbers as explained above.

5.

Columns 5-12 (I(1-8)) correspond to the eight twisted-pair strands within each cable. The cable pairs are counted 1 through 8 from the blue pair on the side of the cable; the pairs are colored blue-gray-gray-gray-blue-gray-gray-gray. Thus the first blue pair counts as 1; the first gray pair next to it is 2; the next gray oaur is 3; the next gray pair is 4; the blue pair in the middle is 5, and so on, as shown in the following table. The entries in the column tell which TSC outputs are connected to each pair. At most two TSCs can be connected to one cable, each TSC is constrained to connect to one half of the cable, and the outputs of the TSC are constrained to connect in decreasing order to the four cable pairs in that half of the cable. The column entries take the form AA-B where AA is the TSC number on that module and B is the output number on that TSC.

Cable to TEC Assignments

At the inputs to the TECs the cells are required to be configured into the proper trigger towers for use by the CFLT[1] in order for subregional energy sums and pattern finding to work correctly. Unfortunately, even with careful assignments of PMTs to FECs and FECs to TSCs, the information is not ordered properly on the eight twisted pair cables to allow a straight mapping of cable ends into TEC connectors. Instead, the cables have to be split down to the single twisted pairs and these pairs have to be mapped onto the TEC front end connector pins. As far as possible FEC and TSC assignments were made to keep one cable feeding into a single connector, or a single half of a crate, but in some cases the cables must even split between crates.

Unlike the other assignments in this series of bringing the photomultiplier signals to the digitization electronics, the assignment of cable twisted pairs to connectors were made by computer. ADAMO relationships were set up from photomultiplier tubes to cables and from photomultiplier tubes to TEC front end connectors, and these relationships were used to map the cable pairs onto the TEC front end connectors.

The connectors have 25 pins each; these are grouped into triplets to form the eight inputs. One pin is left over, pin 13; this pin is grounded. The grouping of the connector pins into the eight inputs is shown in Table 4 and in Figure 54. Each of the three pins in a triplet is connected to one wire from within a cable pair. The white wire is connected to the positive pin, the grey to the negative pin, and the bare wire is connected to ground.

On the front of each TEC there are two such 25 pin connectors, designated connector A and connector B. Connector A, the upper connector, is meant to take most of the inputs. Connector B is only used when sums have to be made at the front end of the TEC.

For the ADAMO relationship of photomultiplier tubes to cable pairs, the photomultiplier tubes were related first to TSCs and then through the TSCs to the cables and cable pairs. In the other direction, the PMTs were related to the trigger towers and through the trigger towers to the TEC front end inputs. In the event that two signals had to be summed at the TEC front end, after the first cable pair was mapped onto the TEC front end connector A, a second cable pair could be mapped onto the second TEC front end connector B. No more than one cable pair could be mapped to one front end connector. If a third cable pair appeared which mapped to the same TEC channel, the program was designed to stop and generate a severe error, since no more than two signals can be summed at the TEC front end. Not all cable pairs map to a TEC connector, as some cable pairs are not connected to TSCs and some are connected to disconnected outputs on TSCs.

TEC Connector Pins
INPUT	PIN NUMBER	CONNECTED TO:
	1	+ (white)
1	2	- (grey)
	14	ground (bare)
	15	+ (white)
2	16	- (grey)
	3	ground (bare)
	4	+ (white)
3	5	- (grey)
	17	ground (bare)
	18	+ (white)
4	19	- (grey)
	6	ground (bare)
	7	+ (white)
5	8	- (grey)
	20	ground (bare)
	21	+ (white)
6	22	- (grey)
	9	ground (bare)
	10	+ (white)
7	11	- (grey)
	23	ground (bare)
	24	+ (white)
8	25	- (grey)
	12	ground (bare)
---	13	disconnected

Table 4. : Pin assignments of the TEC front connectors.

Using ADAMO to Store Cable Pair to TEC Connector Assignments

The cable pair to connector assignments set up through the relations briefly described above are entered into an Adamo table. The entries to this table are made purely by the computer, with checks for possible double entries, such as one cable pair connecting to two different TEC inputs. Two tables are generated, one for the ``A'' connector on the TECs and one for the ``B'' connector, which takes extra signals which must be summed at the TEC front end. As a general rule, all ``A'' connector pins should have cables connected to them, but many ``B'' connector pins do not. The table for the ``A'' connectors is called ``TecConA''; the table for the ``B'' connectors is called ``TecConB''. Both tables have exactly the same format; a description of this format follows. The tables are given in Appendix D.

Each line in the tables corresponds to a TEC. The columns give the following information about the TECs:

• Column 1 (Region) gives the CFLT calorimeter region for which this TEC is designated. Each calorimeter region corresponds to a crate of trigger electronics, so the calorimeter region number also designates the crate in which that TEC sits. The regions are numbered 0-F. Regions 0-3 correspond to the FCAL, regions 4-7 are in the RCAL, and regions 8-F are in the BCAL. The positions of the crates in the Rucksack are shown in Figure 55.
• Column 2 (Card) gives the card number. There are 14 TECs in the crates and split J2 and J3 backplanes. The backplane halves are labelled left and right. Seven TECs sit on each half backplane, and they are numbered 0 through 6 on that backplane. The numbers appearing in this column represent the TEC's number within the half backplane so they run from 0 to 6.
• Column 3 (Bpl) tells upon which backplane the TEC sits. This can be either LEFT or RIGH(t).
• Columns 4-13 (I(1-8)) stand for Inputs(1-8) on a TEC connector, numbered from top to bottom. These are the inputs on the TEC described in Table 4. The entries in these columns give the twisted-pair cable pair which should be connected to this input. They take the form ABB-CC-D where A is F, R, or B (for FCAL, RCAL, or BCAL). BB is the module number within the calorimeter section; it runs from 1 to 23 for the FCAL and RCAL and from 1 to 32 for the BCAL. CC gives the cable number within that module; this can run from 4B to 6B and 4T to 6T. D gives the number of the cable pair; these are numbered 1 through 8 starting with the outside blue pair at the top. Each cable pair contains three wires; these are connected to the three pins making up one input to the TEC according to Table 4. The cable numbering here duplicates that explained for the Cable table previously.

A third table exists which is the inverse of the above table. Instead of listing the cable pairs connected to each connector input, it lists the connector inputs to which the pairs of a cable are attached. It is made in exactly the same manner as the above table, using the same relationships. This cable is named CableSplit and contains information in the following format:

• Column 1 (CAL) gives the calorimeter section from which the cable comes. It takes the values FCAL, RCAL, and BCAL.
• Column 2 (Modul) gives the module within that calorimeter from which that cable comes. Its values are 1-23 in the FCAL and RCAL and 1-32 in the BCAL.
• Column 3 (Cabl) gives the cable number within that module. These cable numbers are the same as those described in the cable table.
• Column 4 (Type) gives the cable type. Again, this is the same as for the cable table.
• Columns 5-13 (I(1-8)) give the destinations for cable pairs 1 through 8. Again, the cable pairs are numbered starting with 1 at the blue pair on one side of the cable. The information on the destination is given in the format R-#B-C-IL, where R is the crate number, which can take values from 0 to F, # is the TEC number, which can go from 0 to six, and B is the backplane designation, which takes values L or R (left or right). C gives the connector designation, which is A for the primary connector and B for the connector which takes extra inputs to be summed at the TEC front end. I is the input number on the connector, which goes from 1-8 and represents groups of three pins to which the cable pair is attached. L can be ``T'' or ``B'' indicating if the TEC input I belongs to a top or a bottom TEC connector. These input numbers are described in Table 4.